You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
I have a question regarding the point filtering strategy for flow generation. During training, you filter out points that do not move beyond a certain distance to focus on moving points. However, at inference time, how do you identify the moving points? Since you only predict flows for the moving points during training, could it negatively impact the overall flow generation process if the input points are not filtered at inference time?
Looking forward to your response!
The text was updated successfully, but these errors were encountered:
Hi,
Thanks for your great work on this project!
I have a question regarding the point filtering strategy for flow generation. During training, you filter out points that do not move beyond a certain distance to focus on moving points. However, at inference time, how do you identify the moving points? Since you only predict flows for the moving points during training, could it negatively impact the overall flow generation process if the input points are not filtered at inference time?
Looking forward to your response!
The text was updated successfully, but these errors were encountered: